真的撞墙？OpenAI 的任务对齐主管：2025年AI会在高阶应用突破，但一般人会感觉停滞

刚刚读到OpenAI 的任务对齐主管Joshua Achiam的分享，他谈到：“对于人工智能的下一阶段，它将在大多数人不知道或不关心的长尾高度专业化技术任务方面变得更好，从而造成一种进展停滞不前的错觉。"

大多数人不知道，不关心，就会让人觉得AI发展停滞不前了？这个逻辑成立吗？个人很好奇他所说的 "高度专业化任务 "是什么？还是说OpenAI真的”撞墙“了？

我们来看看这个老哥说了什么？

我预期将会出现一种奇特的现象：在 AI 的下一个阶段，它将在大量高度专业化的技术任务上表现得更出色，而这些任务大多数人并不了解或关心，这将制造出一种进展停滞的错觉

"A strange phenomenon I expect will play out: for the next phase of AI, it’s going to get better at a long tail of highly-specialized technical tasks that most people don’t know or care about, creating an illusion that progress is standing still."

研究人员会达成他们认为极其重要的里程碑，但多数用户在当时并不会理解其重要性。

"Researchers will hit milestones that they recognize as incredibly important, but most users will not understand the significance at the time."

整体的鲁棒性将会逐步增强。一年之内，常用的大模型在编程、写作和日常事务方面的可靠性将显著提高。然而，鲁棒性的提升并不引人注目，因此很多人可能不会察觉到其中的变化。

"Robustness across the board will increase gradually. In a year, common models will be much more reliably good at coding tasks, writing tasks, basic chores, etc. But robustness is not flashy and many people won’t perceive the difference."

或许在某个时刻，也许是两年之后，人们会注意到 AI 已经深度融入了几乎所有的商业领域，因为它已经跨越了所有的可靠性门槛。这就像智能手机从 2007 年的新奇事物演变为 2010 年代的无处不在一样。

"At some point, maybe two years from now, people will look around and notice that AI is firmly embedded into nearly every facet of commerce because it will have crossed all the reliability thresholds. Like when smartphones went from a novelty in 2007 to ubiquitous in the 2010s."

对于接下来会发生的事情，我很难预测。许多因素存在不确定性，且依赖于未来的发展路径。我唯一有信心的预测是，到 2026 年，Gary Marcus(一直以来对深度学习持怀疑态度的知名学者) 将会坚持认为深度学习已经触及瓶颈

"It feels very hard to guess what happens after that. Much is uncertain and path dependent. My only confident prediction is that in 2026 Gary Marcus will insist that deep learning has hit a wall."

（附录：整个讨论甚至算不上什么预测。这大致反映了自 2023 年初 GPT-4 发布以来的讨论现状，并预计这一趋势将会持续。各种改进和突破的长尾效应正被严重低估）

"(Addendum: this whole thread isn’t even much of a prediction. This is roughly how discourse has played out since GPT-4 was released in early 2023, and an expectation that the trend will continue. The long tail of improvements and breakthroughs is flying way under the radar.)"

我个人倾向于OpenAI内部GPT-5训练可能真的不顺，一个真正的好东西大众却无法感受，这个逻辑上不成立，不过据说o1正式版马上就要发布了，大家对o1的预览版推出时是什么感受？有没有达到你的预期？Sam Altman 之前有说o1正式版要比预览版强很多，果真如他所说，那2025年还真的值得期待的